Abstract :
[en] This doctoral dissertation contributes to the modeling of risk and expected returns in the fields of finance and macroeconomics. On the finance side, the thesis proposes a portfolio choice model and a risk-return setup that simultaneously deal with agents’ unknown utility function, incomplete knowledge of financial return distributions, departures from the Gaussian distribution (i.e. asymmetry and fat-tail risks), investment horizons and investors’ objectives in terms of expected returns. On the macroeconomic side, the objective is to deliver a comprehensive picture of the financial sector in a general equilibrium framework, which accommodates the heterogeneity of behaviors within financial intermediaries.
The first part of the thesis introduces an innovative risk measure, risk horizon, with reference to the speed of convergence of an asset’s mean return to its expectation. This measure is a keystone to a general framework for characterizing investors’ behavior in portfolio selection, which takes into account consideration for volatility, asymmetric and fat-tail risk, a trade-off between downside and upside potential of financial assets, and the timing and probability of deviations from expected returns. The risk horizon framework opens up the way to the identification of forward-looking determinants of market sentiment that includes, among others, the expected market and credit returns.
The last part of the thesis is devoted to a macroeconomic model with heterogeneous and financially constrained intermediaries. An analysis of endogenous risk mechanisms when traditional and shadow banking interact is carried out. The model sheds light on the importance of relative leverage behaviors in the amplification of adverse shocks in the economy.